Contents lists available at ScienceDirect

Urban Forestry & Urban Greening

journal homepage: www.elsevier.com/locate/ufug

Manipulating hedgerow quality: Embankment size influences animalbiodiversity in a peri-urban context

S. Lecqa,⁎, A. Loisela, S.J. Mullinb, X. Bonneta

a Centre d’Etudes Biologiques de Chizé, CEBC CNRS UPR 1934, 79360 Villiers en Bois, FrancebDepartment of Biology, Stephen F. Austin State University, Nacogdoches, TX 75962, USA

A R T I C L E I N F O

Keywords:Non-lethal methodRapid biodiversity assessmentRefugeSpecies richness

A B S T R A C T

Hedgerowsare important features within urban, peri-urban, and agricultural habitats because they shelter mostof the biodiversity in a landscape dominated by infrastructures or a monoculture. Hedges are characterized bytheir vegetative cover but also by their base, notably the breadth of the embankment and the various micro-habitats made by stones, coarse woody debris, and leaf litter. These features determine the availabilities ofarboreal and ground refuges. Their respective roles on biodiversity remain poorly explored. We experimentallymanipulated the size of the embankment in newly-constructed hedges in a peri-urban context. We used non-lethal rapid biodiversity assessments and functional indices (accounting for body mass, trophic level, and me-tabolic mode) to monitor the presence of a wide range of animal taxa. We observed a positive effect of em-bankment size on animal biodiversity. Various elements of the fauna (e.g. arthropods, reptiles) rapidly colonizednewly-constructed hedges provided with an embankment. Guidelines to restore hedgerows should considerembankment size and quality. Both of these features can be improved by simply retaining the materials that areextracted when establishing agricultural plots such that a diversity of microhabitats and ground refuges becomeavailable.

1. Introduction

Urban sprawl and transport infrastructure expansion are leadingcauses of forest fragmentation and habitat alteration, and the con-comitant loss of biodiversity (Wilcox and Murphy, 1985; Savard et al.,2000; Seto et al., 2012). Furthermore, conversion of forest habitats toagricultural use has yielded more than 1.5 billion ha that are currentlycultivated, representing> 10% of the surface of the planet and morethan 36% of the land surface (Bruinsma, 2003). It has been estimatedthat an additional 2.7 billion ha of forests might be progressively con-verted for crop production in the coming decades (Van Vliet et al.,2017).

Certain types of anthropogenic modifications of the landscape canbe beneficial to the wildlife (Fahrig et al., 2011; Pe’er et al., 2014).Natural or managed forests offer refuges for many organisms in highlyaltered urban and agricultural landscapes (Savard et al., 2000; Alvey,2006). Yet the space available is strongly constrained by infrastructures(buildings, roads, etc.). Many urban forests are linear, bordering roads,parks or rivers (Faiers and Bailey, 2005). The benefits of urban foreststo wildlife inhabitants depend on the connectivity among patches;corridors shelter more biodiversity compared to isolated parcels

(Mörtberg and Wallentinus, 2000). Linear forests provide essentialsystems of exchange between peri-urban areas and inner zones of cities,especially alongside rivers and railways (Varet et al., 2013). The ben-efits for biodiversity and human welfare that stem from promotingurban forest networks connected to surrounding habitats are now im-plemented into urban planning strategies (Goddard et al., 2010).

Hedgerows (i.e. linear forests) shelter most of the biodiversity inagricultural and urban landscapes, and they contribute to spatial andstructural heterogeneity (Burel, 1996). Trees are the most salient part ofhedgerows, but previous investigations of the value of this habitat alsoconsidered bordering herbaceous strips and connectivity with otherhabitats (Hinsley and Bellamy, 2000; Moonen and Marshall, 2001;Bailey, 2007). Little attention has been paid to the base of the hedgeshowever, especially the embankment: stones, coarse woody debris, treeroots form a complex matrix of burrows and refuges (Lecq et al., 2017).These structures offer microhabitats for a wide range of organisms andsubstantially contribute to species richness (Lecq et al., 2017). More-over, complex interactions exist among species and many animalsroutinely shuttle between ground shelters and the tree cover above(Ctifl, 2000). Unfortunately, the parameters of the embankment aretypically not accounted for in planting or management guidelines for

https://doi.org/10.1016/j.ufug.2018.08.002Received 20 December 2017; Received in revised form 24 April 2018; Accepted 5 August 2018

⁎ Corresponding author.E-mail address: stephane.lecq@gmail.com (S. Lecq).

Urban Forestry & Urban Greening 35 (2018) 1–7

Available online 07 August 20181618-8667/ © 2018 Published by Elsevier GmbH.

T

hedgerows.We focused on the embankments of hedgerows. Large trees have

broad bases that provide abundant ground refuges and, as such, thecontributions of tree cover to the biodiversity of a given hedgerow arenot easily dissociated from embankment size. For this reason, our ex-periment manipulated the basal structure of hedges independently fromvegetative cover. Hedgerows were installed in a meadow connected toan agricultural landscape and a small city. Three types of embankmentswere created and we sampled animal biodiversity during two years. Toexamine the influence of embankment quality on biodiversity, we useda non-lethal survey method to limit the impact of sampling on ourdependent variable, both for ethical reasons and to encompass a widerange of taxa (i.e., from small invertebrates to mammals; Lecq et al.,2015).

The primary goal of our study was to examine specific features thatcan improve hedgerow quality with minimal impacts on managers andfarmers in terms of costs and labor. We addressed two questions: (1)Does the presence of an embankment promote animal biodiversityfollowing the installation of a hedgerow? (2) Does embankment sizecontribute to that measure of diversity?

2. Material and methods

2.1. Study site

The experiment took place in western central France (46°07′21″N,0°21′24″W) in a typical peri-urban landscape that has been modifiedextensively in the past several decades (Baudry and Jouin, 2003).Traditional agriculture has been replaced by intensive practices whileurbanization developed along with an expanding infrastructure. At thestudy site, approximately 50% of the hedges have been removed (Fig. 1)and most residential areas were constructed recently (e.g., within thelast 25 years; almost all properties containing dwellings that are visiblein Fig. 1).

2.2. Experimental design

In February 2011, nine hedges were created in a rectangular grassmeadow (115× 80m) with the main axis oriented ∼30 °N. Thesouthern margin of the meadow was bordered with a 900m-long hedge(35m width) connected to a village in the east and to a forest in thewest (600m distant). The northern margin was adjacent to a workingarea used for gravel storage and to a road. The two other margins wereadjacent to a cultivated meadow (west) and to an athletic field (east).Prior the experiment, the meadow was regularly mowed and no trees orbushes were present.

The nine hedges were oriented west-east in order to present one sideto maximal sun exposure. They were regularly spaced (10m) and eachmeasured 60m in length. Each hedge was planted with 61 young trees(< 1.5m in total height) representing species that occur locally (e.g.Carpinus betulus, Corylus avellana, Prunus spinosa, Crataegus monogyna inalternation). We constructed each hedge using one of three types ofembankment size: minimal, small and large. The minimum base (MB;n=3 hedges) lacked any sort of embankment; thus, the surface waslevel with the existing grade. Each tree was planted directly in theground. Each hedge having a small base (SB; n= 3 hedges) included asmall (1.00 m wide, 0.75m high) embankment constructed using earthand small stones. We planted the trees on the top of the ridge formed bythe embankment. Hedges having a large base (LB; n=3 hedges) dif-fered only in the size of the embankment (1.50× 1.20m), and treeswere again planted at the top of the embankment ridge. The volume ofmaterial used to construct the LB hedgerows was twofold greater thanthe amount needed for the SB type. In addition, we placed severalstones (∼40× 40 cm) on the south slope in order to cover ∼5% of theground surface of the LB hedges. The embankments were not com-pacted. The three types of embankment correspond to the most

widespread and traditional basal structures found among the hedge-rows in the study area. Overall, we adopted a simple and realistic ap-proach by selecting easily-built structures. In practice, the dimensionsof the SB and LB base types corresponded to the amount of material thatcan be excavated with a backhoe loader during replanting programs. Inareas of our study site where the arable soil is relatively thin, the un-derlying marly-calcareous layers must be broken up before plantation.The three types of hedges were placed in semi-random order, avoiding aconfiguration that placed two of the same hedge type next to each other(Fig. 2). This design enabled us to focus on the effect of the size of theembankment.

Following the construction of the hedges (February 2011), the areawas not managed nor did we monitor the herbaceous vegetation. Thepurpose was to monitor colonization of the hedges by various animalspecies. The proximity of the forest, and the connection of the meadowwith large hedges, provided a putative means by which non-flyingspecies could colonize the hedges within the meadow (Alignier andDeconchat, 2013). For example, many organisms such as woodlouses,myriapods, cryptic spiders, reptiles or small mammals avoid crossingopen areas and follow corridors. To contrast the constructed hedgerowswith those bordering the meadow, the former type is hereinafter termedthe experimental hedges.

2.3. Biodiversity sampling

Many studies addressing the ecological impacts of agriculturalpractice have successfully used birds as an index of animal biodiversity(e.g., Pe’er et al., 2014). Others have focused on a particular taxon (e.g.,Cole et al., 2002). Approaches that integrate a more accurate ex-amination of biodiversity are preferred, however, because using only afew taxa as surrogates of overall diversity provides unreliable assess-ment that might not reflect all ecological processes occurring in thathabitat (Van Jaarsveld et al., 1998; Andelman and Fagan, 2000;Verissimo et al., 2011). Therefore, we did not focus on a given taxo-nomic group and instead attempted to sample all macroscopic animals.

The macrofauna was sampled using five complementary versions ofa protocol developed to visually identify morphospecies: non-lethalrapid biodiversity assessment (NL-RBA; Lecq et al., 2015; Książkiewicz-Parulska and Gołdyn, 2017). Individuals were not collected, but di-rectly identified and/or photographed in the field. Precise speciesidentification was not always possible; instead observed specimenswere assigned to different taxonomic categories, from the species- (fineresolution) to the order-level (coarse resolution; Oliver and Beattie,1996). To limit observer bias, pictures of common and difficult specieswere used as “reference specimens;” moreover pictures were takenrandomly or when the observer was uncertain. This limitation of theNL-RBA approach was offset by the absence of environmental or ethicalconcerns, the ability to sample a wide range of taxa, and a high cost/efficiency ratio that enabled us to accumulate a large data set (Lecqet al., 2015).

Three versions of the NL-RBA protocol (rapid visual transect, slowvisual transect and focal observation) were relatively similar as theyrelied on visual searching of the fauna using different walking speeds.The two other versions (active searching and cover objects [five cor-rugated slabs of cement were placed along each experimental hedge])targeted cryptic fauna that typically associate with ground refuges.Natural shelters such as stones, leaf litter, or artificial shelters werelifted during these survey variations (see Lecq et al. (2015) for detailson each version of the protocol). These two last survey methods at-tempted to detect hidden animals and, as such, could produce an en-counter even at times outside of the activity period for a given species.By combining the different versions of the NL-RBA, the methods weredesigned to include the relatively cryptic species that depend on theavailability of ground refuges (e.g. arthropods, mollusks, reptiles). Yet,many individuals belonging to not-cryptic species were also counted(e.g., a pair of wagtails successfully nested in one experimental hedge).

S. Lecq et al. Urban Forestry & Urban Greening 35 (2018) 1–7

2

Fig. 1. Aerial view of the broad context of the study (Google-Earth 2015), situated within in a typical peri-urban landscape where agricultural fields are partlybordered by hedgerows. The city of Chizé is in the right side of the image. The white square indicates the enlarged portion of the Fig. 2. The dashed white lines showthe past position of hedgerows destroyed during the last decades. Likely, a greater number of hedges have been removed than indicated.

Fig. 2. Aerial view of the experimental studysite. Nine hedges (black lines) were built in ameadow (white rectangle). The enlargementshows the three types of experimental hedges:short dashed line indicates lack of embank-ment, the medium dashed line indicates smallembankment (1 m wide × 0.75m high), andthe continuous line shows large embankment(1.50× 1.20m). The southern limit of themeadow was adjacent to an existing hedgerow.A gravel pit, a hedge and a road form thenorthern limit. The other two borders wereadjacent to relatively open habitat, notably astadium on the east side (see text for details).Nearest estates are less than 300m away.

S. Lecq et al. Urban Forestry & Urban Greening 35 (2018) 1–7

3

For simplicity, each occasion that a hedge was surveyed was called amonitoring session regardless of the protocol used.

The monitoring period extended from spring 2011 (April) tosummer 2012 (August), but winter (October–March) and drought per-iods (in July) were excluded. The experimental hedges were colonizedrelatively quickly (e.g., after only 2 month of maturation). Each of thefive versions of the protocol was performed 5 to 8 times in each ex-perimental hedge. A total of 328 transects was surveyed, and each ex-perimental hedge was surveyed an average (± 1 SD) of 7.3 ± 0.8times. The different types of transects were equally distributed acrosseach hedge (Lecq et al., 2017).

2.4. Analyses

All observed specimens were assigned to a morphospecies. Wecompared the number of morphospecies between the different types ofexperimental hedges (a proxy of the diversity of morphospecies thatcolonized the hedges). Individuals were not marked (and not killed),and thus some of them might have been observed more than once. Asingle individual counted more than once, however, cannot belong tomore than one morphospecies and, as such, our analyses were notsubject to pseudoreplication.

A limitation to using the number of morphospecies as a metric ofdiversity is that the ecological roles of each morphospecies are notidentical. For example, a large predator (> 100 g) should not be di-rectly compared to a small primary consumer (1 g) because the formerdepends on a more complex trophic chain, and therefore a largeramount of resources, than the latter. Our data set includes such a widerange of morphospecies (e.g. large colubrid snakes versus grasshoppers).In order to account for the respective weight of the different mor-phospecies observed, we used a simple functional index that accountsfor body mass, trophic level, and metabolic mode (Lecq et al., 2017).The criteria used in this index to characterize, and thus rank the mor-phospecies, were conservative. For each morphospecies the index wascalculated using the following equation:

Ecological rank= log(body mass)× trophic level×metabolic mode

Details for the estimation of the parameters are provided in Lecqet al. (2017). As an example, the shift from ectothermy to endothermy(metabolic mode) was associated with a factor of 2 even though, onaverage, endothermic animals exhibit a metabolism 10 times greatercompared to ectothermic species.

For each hedgerow, we used the sum of the ecological ranks of thedifferent morphospecies observed divided by the number of monitoringsessions. The value obtained was named ‘hedgerow biodiversity score’(HB). This score differed from the numbers of morphospecies observedin the hedges – it provided an index of the complexity of the trophicchain instead of an index of morphospecies richness.

Predators depend on underlying trophic levels and members of thisgroup are considered as a useful surrogate for biodiversity (Sergio et al.,2009). For this reason, we compared the proportion of predator mor-phospecies to the total number of morphospecies. In our sample, mostpredators were represented by various species of invertebrates (e.g.spiders, carabids). Using NL-RBA, predators cannot be confused withother consumers: for example, even if identified incorrectly at thespecies level, a spider will be correctly assigned to spiders. For thisspecific analysis, possible pseudoreplicates were included. However,considering the duration of the monitoring period (1.5 year) and be-cause the time elapsed between sessions was usually greater than onemonth, the likelihood of counting the same individual several times waslimited. Moreover, the effect of pseudoreplicated data should applyequally among the three hedges types and thus did not impede com-parisons among them.

Our analyses also considered the number of reptiles. Three species

were observed in the experimental hedges (two colubrid snakes and onelacertid lizard) and individuals were individually marked (enabling usto discard possible pseudo-replicates). They are all predators, but thetwo snakes (Hierophis viridiflavus and Zamenis longissimus) are amongthe largest predatory species (> 1.3 m,>400 g) in the region. Theyfeed on vertebrates and their diet includes other predatory species (e.g.lizards, shrews, birds; Lelièvre et al., 2012). They exhibit sedentaryhabits and they depend on ground refuges (Bonnet et al., 1999; Lelièvreet al., 2010) especially in hedgerow systems (Bonnet et al., 2016; Lecqet al., 2017). More generally, the herpetofauna offers suitable surrogatetaxa to estimate local biodiversity, and as such they are considered asuseful bio-indicators (Beaupre and Douglas, 2009; Lewandowski et al.,2010). Although reptiles were included in the two other analyses (i.e.,numbers of observations and HB score), their contribution was limited(in reptiles, N= 23 observations and N=3morphospecies, comparedto the arthropods where N > 7000 observations and N > 200morphospecies). Thus, the use of reptile number was considered as arelatively independent index.

The distribution of the variables tested did not deviate from nor-mality (e.g. HB, Shapiro-Wilk normality test W=0.98, P= 0.26), thuswe used analyses of variance (ANOVA) for our comparisons. None ofthe hedges provided disproportionate amount or lack of data; insteadeach set of three hedges delivered relatively homogenous signal as in-dicated by the relatively modest standard deviations (see Fig. 3; nooutlier effect). Proportional values were arsine-transformed prior ana-lyses.

3. Results

3.1. General trend

We obtained a total of 13,776 observations during the study periodthat comprised a total of 237 morphospecies (Table 1). The resultsconsistently indicated a positive effect of embankment size on the threeindices (Fig. 3). Thus, the shelters provided by the embankment per se,and by the greater amounts of refuges (e.g. stones) influenced thelikelihood of observing a greater number of morphospecies, notablylarge predators.

3.2. Number of morphospecies

Embankment size influenced the mean number of morphospeciesobserved per hedge (ANOVA, F2,6= 6.71, P < 0.03; Fig. 3). A Kruskal-Wallis test returned a similar result (H=5.96, P=0.05). Post-hoc testsindicated a difference in the number of morphospecies between onlythe MB and LB embankment types (Bonferroni post-hoc test, P= 0.03;Fig. 3).

3.3. Hedgerow biodiversity score (HB)

The mean HB score differed among the three types of experimentalhedges (ANOVA, F2, 6= 72.98, P < 0.001; Kruskal-Wallis H=7.20,P= 0.03; Fig. 3). Bonferroni post-hoc tests indicated differences in HBscores among all hedgerow types (0.010 < P < 0.001).

3.4. Proportion of predator morphospecies

The proportion of predatory species differed across the three hedgetypes (ANOVA, F2, 6= 28.35, P < 0.001; Kruskal-Wallis H=5.96,P= 0.05; Fig. 3). These values were greater in the SB and LB hedgetypes than in the hedges without any embankment (Bonferroni post-hoctest P < 0.01). However, differences in this response variable were notapparent between SB and LB hedge types (P=0.27).

S. Lecq et al. Urban Forestry & Urban Greening 35 (2018) 1–7

4

3.5. Number of reptiles

We found no reptile in the experimental hedge with no embank-ment. The mean number of reptiles increased with embankment size(ANOVA: F2, 6= 42.25, P < 0.001; Kruskal-Wallis H= 9.00,P=0.01; Fig. 3).

4. Discussion

To our knowledge, this study is the first to focus specifically on themicrohabitats at the basal areas of hedges: the embankments. Althoughframed in a peri-urban context, this study has broader implications.Large amounts of hedgerow habitat have been removed from agri-cultural areas during the past several decades leading to a markedhomogenization of the landscapes, especially in Europe (e.g., Baudryand Jouin, 2003). Detrimental effects on biodiversity, with a noticeableloss of ecosystem services, have been observed (Le Coeur et al., 2002;Batáry et al., 2015). This study examines the relative importance oflocal management in order to propose cost effective recommendations

to limit the biodiversity loss caused by landscape modifications. Theconsistent signal across all response variables was that changes in thebasal structure of the hedge influenced the occurrence of morphos-pecies. Indeed, each of the different indices for the presence of mac-roscopic fauna exhibited a positive relationship with the size of theembankment base. This result was expected because many animalspecies depend on the presence of appropriate refuges, notably crypticpredators belonging to various taxa (e.g. arachnids, carabids, reptiles).Furthermore, the positive relationship between the occurrence of theembankments and the diversity of morphospecies could be driven bymore abundant or higher-quality ecotone habitat being available (Pereaet al., 2011; Klar and Crowley, 2012). We next review the mainmethodological limits of this study, and then discuss results and per-spectives.

The main limitation of our study is represented by the relativelysmall number of experimental hedges per treatment (e.g. 3 hedges persize-group, 5 added cover objects per hedge). Despite this small numberof replicates, confidence intervals remained narrow (Fig. 3). Thus, theconsistency within groups was associated with the notable differences

Fig. 3. The effect of different embankments ondifferent indices of macrofaunal diversity inhedgerows: no embankment (= ‘minimal’),small, and large embankments (see text fordetails). The number of morphospecies re-presents various taxa of invertebrates and ver-tebrates. HB scores account for the size, dietand metabolic mode of the morphospecies.Values are shown as means +1 standard de-viation. Letters above bars denote significantdifferences (post hoc tests) among hedge types.

Table 1Number and occurrences of morpho-species observed in the hedgerows.

Morpho-species Occurrence

Large Base Small Base Minimum Base Total Large Base Small Base Minimum Base Total

Arachnida 30 34 23 40 523 495 315 1333Aves 1 1 1 2 5 16 11 32Chilopoda 3 3 1 4 6 4 1 11Clitellata 1 0 1 1 1 0 2 3Crustacean 1 1 1 1 317 256 232 805Diplopoda 1 0 0 1 2 0 0 2Gastropoda 8 7 7 8 2185 2525 1782 6492Insecta 124 111 106 171 2000 1756 1178 4934Mammalia 2 3 5 6 20 18 39 77Reptilia 3 1 0 3 15 8 0 eNothing NA NA NA NA 5 6 48 59Total 174 161 145 237 5079 5084 3608 13771

S. Lecq et al. Urban Forestry & Urban Greening 35 (2018) 1–7

5

among groups, suggesting that our results were robust. A possible lackof independence among the hedges attributable to the short distancebetween them did not influence the interpretation of our results. Duringcolonization, individuals had the capacity to navigate across themeadow. Indeed, with the exception of sessile life-history stages (e.g.lepidopteran chrysalis) almost all the morphospecies studied had thephysical ability to cross the distance between adjacent hedges (10m).This means that they had ample opportunities to select settlement orforaging places, especially considering the long time periods elapsedbetween consecutive sampling events. Thus, a greater number of mor-phospecies (notably, large predators) actually preferred hedges with alarge embankment, where shelter availability was higher.

The relative homogeneity of structure within each of the threehedge types (i.e., similar age and species composition of planted trees)indicates that embankment size was more important than the exactposition of the hedges within the meadow. Finally, limitations asso-ciated with the use of slightly different NL-RBA protocols were offset byadvantages. Our approach incorporating multiple survey modes en-abled us to sample a wide range of taxa, and to collect a large data set ina non-destructive way. Lethal methods remove important numbers ofindividuals and might hinder establishment in newly-constructed ha-bitat, especially for species or populations that are sensitive to periodictakings. In practice, identification errors were limited to those speciesthat are morphologically similar in size and in shape. Further, three ofthe four indices do not rely on taxonomic accuracy and did not presentmajor identification errors (Lecq et al., 2015; Książkiewicz-Parulskaand Gołdyn, 2017).

Overall, our conclusions that increasing embankment size promotesanimal biodiversity, augments the proportion of predators, and createsattractive conditions for reptiles are likely robust. During the studyperiod, the planted trees played a minor role (if any) because they hadinsufficient time for any appreciable above-ground growth.Precipitation levels were low during the experiment (MeteoFrance) andalmost all trees remained small (< 50 cm in height). Continued growthof the trees and shrubs, and the associated production of leaf litter,should further enhance biodiversity on the local scale (Zanaboni andLorenzoni, 1989). Long-term experiments are required to examine thisissue, however, because the typical vegetation along hedgerows re-quires at least 20 years to reach maturity.

Reptiles were observed only in the hedges provided with an em-bankment (SB and LB treatments). Lizards and snakes use ground re-fuges intensively, and thus were likely attracted by the shelters offeredby the embankment and stones (Bonnet et al., 1999, 2013). Our ex-perimental hedges were built in a set-aside meadow adjacent to a largehedgerow (Fig. 2). Reptiles were probably present in nearby habitats atthe beginning of the study. The rapid colonization of the hedges wasenhanced by the connectivity with surrounding favorable habitats (e.g.large hedge, forest) that represented a source population of individuals.Thus, our results might not be transposable to large industrial and in-tensively managed urban-ecosystems devoid of source of colonizers.This does not diminish the potential importance of embankments forthose species that can cross large distances, notably aerial dispersers orflying species. Hedgerows with large embankments might be useful toreinforce the corridors and connection among structures (Mauremootoet al., 1995) and to promote the other beneficial roles of hedgerows(Xiao et al., 2010). Regarding reptiles specifically, a long-term experi-ment set up in a peri-urban park demonstrated that promoting snakepopulations through habitat management was possible and well ac-cepted by residents (Bonnet et al., 2016). The initial conditions in thatstudy were particularly challenging for several reasons: 1) hedgerowswere created through tree-cutting (people prefer large trees overshrubs), 2) the main snake species involved (Vipera aspis) was veno-mous, potentially dangerous, and thus undesired in public areas, and 3)the park was amidst a vast urban area and was thus relatively isolatedfrom other natural habitats. The current experiment did not involvethese degrees of difficulties. Convincing managers and people to accept

and to promote complex structures at the basement of hedgerows (e.g.large embankments) that might potentially attract small invertebratesand reptiles should represent an easier objective.

Our study provides information that offers practical outcomes forhabitat management and restoration. During planting programs, forexample, recommendations can be proposed to improve the biodi-versity and carrying capacity of hedges. For aesthetic reasons, materialexcavated from a site is typically deposited elsewhere. We suggest thatretaining this material on site to build uncompacted embankments willcreate microhabitats that are attractive to those organisms that aredesired - notably predators of plant pest species like parasitoids (Ctifl,2000; Langer, 2001; Tscharntke et al., 2005; Batáry et al., 2010). Spe-cific educational programs might be useful. The relatively rapid colo-nization (< 1.5 year) by more than 200 different morphospecies sug-gests that including embankment structures in construction guidelinesfor hedgerows might present a cost-effective method that favors animalbiodiversity. Although not investigated in this study, plant biodiversitymight also respond positively to the presence of embankments, a pro-cess that could further increase animal diversity (Lawton, 1983).

5. Conclusions

This study provides an experimental support to a recent empiricalinvestigation that suggested a positive influence of embankment size onanimal biodiversity in agricultural landscapes (Lecq et al., 2017). Ourstudy design involved the construction of realistic embankments - thethree types of experimental hedges used were representative of the si-tuations encountered in peri-urban areas and agricultural landscape ofthe region in which we conducted the study. It would be interesting toreplicate this experiment in different contexts (e.g. parks, rivers, gar-dens). Furthermore, it could be useful to augment features of the hedgewith shelters (like stone heaps) or plant debris to further bolster bio-diversity. The loss of ecological heterogeneity at multiple spatial andtemporal scales is a universal consequence of the intensification ofanthropogenic activities. Future research should develop integrativepolicy frameworks and management solutions that re-establish thatheterogeneity. Implementing habitats and microhabitats into largescale restoration programs to promote the conservation of biodiversityhas been identified as a priority (Gren et al., 2014; Crouzeilles et al.,2015). Practical solutions to reach these objectives, one of which hasbeen examined in this study, should imply a high efficiency/cost ratioto be applicable.

Acknowledgements

The city of Chizé (France 79) generously provided the land for theexperimental meadow. We warmly thank the students, notably J.Souchet, and volunteers who assisted with the NL-RBA. We also thankthe specialists, notably B. Leroy and J. Yvernault who assisted withspecies identification. Two anonymous reviewers provided constructivecomments.

References

Alignier, A., Deconchat, M., 2013. Patterns of forest vegetation responses to edge effect asrevealed by a continuous approach. Ann. For. Sci. 70, 601–609.

Alvey, A.A., 2006. Promoting and preserving biodiversity in the urban forest. Urban For.Urban Green. 5, 195–201.

Andelman, S.J., Fagan, W.F., 2000. Umbrellas and flagships: efficient conservation sur-rogates or expensive mistakes? Proc. Natl. Acad. Sci. 97, 5954–5959.

Bailey, S., 2007. Increasing connectivity in fragmented landscapes: an investigation ofevidence for biodiversity gain in woodlands. For. Ecol. Manage. 238, 7–23.

Batáry, P., Matthiesen, T., Tscharntke, T., 2010. Landscape-moderated importance ofhedges in conserving farmland bird diversity of organic vs. conventional croplandsand grasslands. Biol. Conserv. 143, 2020–2027.

Batáry, P., Dicks, L.V., Kleijn, D., Sutherland, W.J., 2015. The role of agri‐environmentschemes in conservation and environmental management. Conserv. Biol. 29,1006–1016.

Baudry, J., Jouin, A., 2003. De la haie aux bocages: organisation, dynamique et gestion.

S. Lecq et al. Urban Forestry & Urban Greening 35 (2018) 1–7

6

Institut national de la recherche agronomique Editions, Paris, pp. 435.Beaupre, S.J., Douglas, L.E., 2009. Snakes as indicators and monitors of ecosystem

properties. In: Mullin, S.J., Seigel, R.A. (Eds.), Snakes: Ecology and Conservation.Cornell University Press, pp. 244–261.

Bonnet, X., Naulleau, G., Shine, R., 1999. The dangers of leaving home: dispersal andmortality in snakes. Biol. Conserv. 89, 39–50.

Bonnet, X., Fizesan, A., Michel, C.L., 2013. Shelter availability, stress level and digestiveperformance in the aspic viper. J. Exp. Biol. 216, 815–822.

Bonnet, X., Lecq, S., Lassay, J.L., Ballouard, J.M., Barbraud, C., Souchet, J., Mullin, S.J.,Provost, G., 2016. Forest management bolsters native snake populations in urbanparks. Biol. Conserv. 193, 1–8.

Bruinsma, J., 2003. World Agriculture: Towards 2015/2030: an FAO Perspective.Earthscan.

Burel, F., 1996. Hedgerows and their role in agricultural landscapes. Crit. Rev. Plant Sci.15, 169–190.

Cole, L.J., McCracken, D.I., Dennis, P., Downie, I.S., Griffin, A.L., Foster, G.N., Murphy,K.J., Waterhouse, T., 2002. Relationships between agricultural management andecological groups of ground beetles (Coleoptera: Carabidae) on Scottish farmland.Agric. Ecosyst. Environ. 93, 323–336.

Crouzeilles, R., Beyer, H.L., Mills, M., Grelle, C.E., Possingham, H.P., 2015. Incorporatinghabitat availability into systematic planning for restoration: a species‐specific ap-proach for Atlantic Forest mammals. Divers. Distrib. 21, 1027–1037.

Ctifl, 2000. Les haies composites réservoirs d’auxiliaires. Éditions Centre technique in-terprofessionnel des fruits et légumes, Paris.

Fahrig, L., Baudry, J., Brotons, L., Burel, F.G., Crist, T.O., Fuller, R.J., Sirami, C.,Siriwardena, G.M., Martin, J.L., 2011. Functional landscape heterogeneity and an-imal biodiversity in agricultural landscapes. Ecol. Lett. 14, 101–112.

Faiers, A., Bailey, A., 2005. Evaluating canal side hedgerows to determine future inter-ventions. J. Environ. Manage. 74, 71–78.

Goddard, M.A., Dougill, A.J., Benton, T.G., 2010. Scaling up from gardens: biodiversityconservation in urban environments. Trends Ecol. Evol. 25, 90–98.

Gren, M., Baxter, P., Mikusinski, G., Possingham, H., 2014. Cost-effective biodiversityrestoration with uncertain growth in forest habitat quality. J. For. Econ. 20, 77–92.

Hinsley, S.A., Bellamy, P.E., 2000. The influence of hedge structure, management andlandscape context on the value of hedgerows to birds: a review. J. Environ. Manage.60, 33–49.

Klar, N.M., Crowley, P.H., 2012. Shelter availability, occupancy, and residency in size-asymmetric contests between rusty crayfish, Orconectes rusticus. Ethology 118,118–126.

Książkiewicz-Parulska, Z., Gołdyn, B., 2017. Can you count on counting? Retrieving re-liable data from non-lethal monitoring of micro-snails. Perspect. Ecol. Conserv. 15,124–128.

Langer, V., 2001. The potential of leys and short rotation coppice hedges as reservoirs forparasitoids of cereal aphids in organic agriculture. Agric. Ecosyst. Environ. 87, 81–92.

Lawton, J.H., 1983. Plant architecture and the diversity of phytophagous insects. Annu.Rev. Entomol. 28, 23–39.

Le Coeur, D., Baudry, J., Burel, F., Thenail, C., 2002. Why and how we should study fieldboundary biodiversity in an agrarian landscape context. Agric. Ecosyst. Environ. 89,23–40.

Lecq, S., Loisel, A., Bonnet, X., 2015. Non-lethal rapid biodiversity assessment. Ecol.Indic. 58, 214–224.

Lecq, S., Loisel, A., Brischoux, F., Mullin, S.J., Bonnet, X., 2017. Importance of groundrefuges for the biodiversity in agricultural hedgerows. Ecol. Indic. 72, 615–626.

Lelièvre, H., Blouin-Demers, G., Bonnet, X., Lourdais, O., 2010. Thermal benefits of ar-tificial shelters in snakes: a radiotelemetric study of two sympatric colubrids. J.Therm. Biol. 35, 324–331.

Lelièvre, H., Legagneux, P., Blouin-Demers, G., Bonnet, X., Lourdais, O., 2012. Trophicniche overlap in two syntopic colubrid snakes (Hierophis viridiflavus and Zamenislongissimus) with contrasted lifestyles. Amphibia-Reptilia 33, 37–44.

Lewandowski, A.S., Noss, R.F., Parsons, D.R., 2010. The effectiveness of surrogate taxa forthe representation of biodiversity. Conserv. Biol. 24, 1367–1377.

Mauremooto, J.R., Wratten, S.D., Worner, S.P., Fry, G.L.A., 1995. Permeability ofhedgerows to predatory carabid beetles. Agric. Ecosyst. Environ. 52, 141–148.

Moonen, A.C., Marshall, E.J.P., 2001. The influence of sown margin strips, managementand boundary structure on herbaceous field margin vegetation in two neighbouringfarms in southern England. Agric. Ecosyst. Environ. 86, 187–202.

Mörtberg, U., Wallentinus, H.-G., 2000. Red-listed forest bird species in an urban en-vironment – assessment of green space corridors. Landsc. Urban Plan. 50, 215–226.

Oliver, I., Beattie, A.J., 1996. Invertebrate morphospecies as surrogates for species: a casestudy. Conserv. Biol. 10, 99–109.

Pe’er, G., Dicks, L.V., Visconti, P., Arlettaz, R., Báldi, A., Benton, T.G., Collins, S.,Dieterich, M., Gregory, R.D., Hartig, F., Henle, K., Hobson, P.R., Kleijn, D., Neumann,R.K., Robijns, T., Schmidt, J., Shwartz, A., Sutherland, W.J., Turbé, A., Wulf, F., Scott,A.V., 2014. EU agricultural reform fails on biodiversity. Science 344, 1090–1092.

Perea, R., Miguel, A.S., Gil, L., 2011. Acorn dispersal by rodents: the importance of re-dispersal and distance to shelter. Basic Appl. Ecol. 12, 432–439.

Savard, J.P.L., Clergeau, P., Mennechez, G., 2000. Biodiversity concepts and urban eco-systems. Landsc. Urban Plan. 48, 131–142.

Sergio, F., Caro, T., Brown, D., Clucas, B., Hunter, J., Ketchum, J., McHugh, K., Hiraldo,F., 2009. Top predators as conservation tools: ecological rationale, assumptions andefficacy. Annu. Rev. Ecol. Evol. Syst. 39, 1–19.

Seto, K.C., Güneralp, B., Hutyra, L.R., 2012. Global forecasts of urban expansion to 2030and direct impacts on biodiversity and carbon pools. Proc. Natl. Acad. Sci. 109,16083–16088.

Tscharntke, T., Klein, A.M., Kruess, A., Steffan‐Dewenter, I., Thies, C., 2005. Landscapeperspectives on agricultural intensification and biodiversity–ecosystem servicemanagement. Ecol. Lett. 8, 857–874.

Van Jaarsveld, A.S., Freitag, S., Chown, S.L., Muller, C., Koch, S., Hull, H., Bellamy, C.,Krüger, M., Endrödy-Younga, S., Mansell, M.W., Scholtz, C.H., 1998. Biodiversityassessment and conservation strategies. Science 279, 2106–2108.

Van Vliet, J., Eitelberg, D.A., Verburg, P.H., 2017. A global analysis of land take incropland areas and production displacement from urbanization. Glob. Environ.Change 43, 107–115.

Varet, M., Burel, F., Lafage, D., Pétillon, J., 2013. Age-dependent colonization of urbanhabitats: a diachronic approach using carabid beetles and spiders. Anim. Biol. 63,257–269.

Verissimo, D., MacMillan, D.C., Smith, R.J., 2011. Toward a systematic approach foridentifying conservation flagships. Conserv. Lett. 4, 1–8.

Wilcox, B.A., Murphy, D.D., 1985. Conservation strategy: the effects of fragmentation onextinction. Am. Nat. 125, 879–887.

Xiao, B., Wang, Q.H., Wu, J.Y., Huang, C.W., Yu, D.F., 2010. Protective function ofnarrow grass hedges on soil and water loss on sloping croplands in Northern China.Agric. Ecosyst. Environ. 139, 653–664.

Zanaboni, A., Lorenzoni, G.G., 1989. The importance of hedges and relict vegetation inagroecosystems and environment reconstitution. Agric. Ecosyst. Environ. 27,155–161.

S. Lecq et al. Urban Forestry & Urban Greening 35 (2018) 1–7

7